1. Wsprdaemon
  2. Topics

2 questions : .c2 files coming from Red Pitaya and MERGE

Rob Robinett
 

1). Decoding .c2 files.
 According to its help menu, the 'wsprd' command which performs the decodes will accept .c2 files, but I have had no occasion to use them.  You can experiment with the use of .c2 files with wsprd by running it from the command line of your WD machine.  To see how WD runs the decodes of wav files you can run this command while WD is running the decodes in the first 20+ seconds of each 2 minute WSPR cycle, e.g.: 

wsprdaemon@KPH-TC:~$ ps aux | grep "wsprd " | grep -v timeout | grep -v grep
wsprdae+ 1386990 36.3  0.2  20252 16212 ?        RN   11:12   0:07 /home/wsprdaemon/wsprdaemon/bin/wsprd -c -C 1000 -o 4 -d -f 14.0956 230224_1910.wav
wsprdaemon@KPH-TC:~$

Then take that command and modify its arguments for your application, e.g.:

wsprdaemon@KPH-TC:~$ /home/wsprdaemon/wsprdaemon/bin/wsprd -c -C 1000 -o 4 -d -f 14.0956 YOUR_FILE.c2

The WD demon for each configured RECEIVER/BAND spawns a wav file recorder and then polls every 10 seconds for the wav files produced by that recorder to stop growing in size.  Currently only KiwiSDR recording is supported and WD runs the kiwi recorder.py program to create them, but I will soon start debugging support for baseband AUDIO and SDR input devices.  A key requirement is that the recording program must create a series of one minute long wav files where the last audio sample of a wav file is then followed by the next audio sample being found as the first sample of the next wav file. 

My limited understanding of c2 files is that they are frequency domain samples, so I don't understand how one can align c2 files in the same contiguous temporal order with no gaps.  But if one can create a series of c2 files with that temporal alignment, then I can easily incorporate processing of them into WD.

2). I have tried to extensively document the WD configuration options in ~./wsprdaemon/wd_template.conf.  In there you will find an example of a MERGed receiver MERG_0:

declare RECEIVER_LIST=(
        "KIWI_0                  10.11.12.100:8073     AI6VN         CM88mc    NULL"
        "KIWI_1                  10.11.12.101:8073     AI6VN         CM88mc  foobar       DEFAULT:-10,80:-12,30:-8,20:2,15:6"
        "KIWI_2                  10.11.12.102:8073     AI6VN         CM88mc  foobar"
        "AUDIO_0                     localhost:0,0     AI6VN         CM88mc  foobar"      ### The id AUDIO_xxx is special and defines a local audio input devic
        "AUDIO_1                     localhost:1,0     AI6VN         CM88mc  foobar"
        "SDR_0                           RTL-SDR:0     AI6VN         CM88mc  foobar"      ### The id SDR_xxx   is special and defines a local RTL-SDR or other
        "SDR_1                           RTL-SDR:1     AI6VN         CM88mc  foobar"
        "MERG_0         KIWI_1,KIWI2,AUDIO_1,SDR_1     AI6VN         CM88mc  foobar"      ### In a device with the name starting with "MERG", the IP field is a
)

And also how to use it in a schedule  named 'WSPR_SCHEDULE_merged':

declare WSPR_SCHEDULE_merged=(
    "00:00                       MERG_0,630 MERG_0,160"
)

Although I don't show it in that example, a schedule can mix MERGed and regular receivers, e.g.:

declare WSPR_SCHEDULE_merged=(
    "00:00                       MERG_0,630 MERG_0,160 KIWI_0,80"
)

I have taken the liberty of posting this reply to the WD group so that others may benefit from your questions and my response.

73,

Rob


On Fri, Feb 24, 2023 at 3:02 AM Erwin Serle <waterwin2@...> wrote:
Dear Rob,

Hope you have a moment to help me with 2 questions.




1. Can wsprdaemon somehow handle .c2 files coming from Red Pitaya for decoding the 2 minute wspr and fst4w signals contained in it. Or would that be a major effort at the software side for you to make that work.

The file name looks like this :  wspr_10_14095599_2_230224_1030.c2

wspr_<band #>_<mid-frequency-in-Hz>_<Red Pitaya antenna channel>_<date>_<UTC-start-time>.c2

I can define a more useable file name if needed


Inline image



2. Can you tell me where to find the explanation of the MERGE function of wsprdaemon so I can use that to gather all my spots. So far my understanding is that I have to put a file with the decoded spots in a certain format into a specific file location that should match the setup in wsprdaemon.conf.
But my understanding might be completely wrong.


Thank you for your assistance,

Erwin




--
Rob Robinett
AI6VN
mobile: +1 650 218 8896
Gwyn Griffiths
 

Rob, Erwin

A few points on .c2 files as generated by wsprd:

  • They are time domain.
  • They contain 45000 samples decimated from the wav file.
  • The format is IQ as binary float32
  • Within WsprDaemon c2 files are generated when wsprd is run, so your concatenating of wav files before wsprd automatically generates a c2 file of two minutes.
  • c2 files are used within WsprDaemon for the 'c2_FFT' noise estimator.
  • Search for # Filename: c2_noise.py within WD program to see the python code as an example of handing these c2 files
regards

Gwyn G3ZIL
Erwin - PE3ES - F4VTQ
 

Pavel Demin's Red Pitaya multiband wspr program is using the 2 minute long .c2 files to decode the wspr spots with wsprd.

He is using the .c2 files because his FPGA design creates .c2 files and no .wav files

Early 2022 I have already shown that .c2 fills can be picked apart at the minute mark and reassembled / concatenated so this should be helpful to get 5/15/30 minute long .c2 files.
BUT jt9 only handles .wav files and decoding FST4W from the Red Pitaya .c2 files looks to be impossible (?)
Rob Robinett
 

If a wav file is transformed into a c2 file, then I expect that a c2 can be transformed into a wav file.
Of course the resulting wav file will contain only the WSPR band frequencies, but those are all we need.

On Sat, Feb 25, 2023 at 7:15 AM Erwin - PE3ES - F4VTQ via groups.io <waterwin2=yahoo.com@groups.io> wrote:
Pavel Demin's Red Pitaya multiband wspr program is using the 2 minute long .c2 files to decode the wspr spots with wsprd.

He is using the .c2 files because his FPGA design creates .c2 files and no .wav files

Early 2022 I have already shown that .c2 fills can be picked apart at the minute mark and reassembled / concatenated so this should be helpful to get 5/15/30 minute long .c2 files.
BUT jt9 only handles .wav files and decoding FST4W from the Red Pitaya .c2 files looks to be impossible (?)



--
Rob Robinett
AI6VN
mobile: +1 650 218 8896
Erwin - PE3ES - F4VTQ
 

Yes by only looking at every other sample instead of at all samples. 

A test could be to re-run the .c2 files generated by wsprdaemon wsprd and see if these still decode

Interestingly Pavel made some changes to ft8d so it accepts .c2 files for his implementation of ft8 decoding as jt9 only accepts .wav files, that was done some years ago together with Andy K1RA.
I do not know if that already was a version that decoded fst4w as well.

1 - 5 of 5